- Email: [email protected]
Mitogenic activity is increased in the sera of preeclamptic women before delivery
Mitogenic activity is increased in the sera of preeclamptic women before delivery Thomas J. Musci, MD, James M. Roberts, MD, George M. Rodgers, MD, PhD, and Robert N. Taylor, MD, PhD San Francisco, California There is increasing evidence that endothelial cell injury and altered endothelial cell function play an important role in the pathogenesis of preeclampsia. Endothelial cell injury can lead to the secretion of potent mitogens by activating platelets and directly through the increased production of peptide growth factors by endothelial cells themselves. This study was undertaken to test the hypotheses that increased secretion of mitogenic factors is a feature of preeclampsia and that this activity could be detected in the serum of preeclamptic women. Paired serum samples were collected in early labor and again at 24 to 48 hours post partum from term patients with preeclampsia (n = 15) and normal pregnant controls (n = 14). A bioassay was used to quantify mitogenic activity in these paired samples by assessing their ability to stimulate the incorporation of tritiated thymidine into deoxyribonucleic acid of confluent, quiescent (G 0 stage) human fibroblasts in monolayer culture. Mitogenic activity was significantly increased in prepartum, preeclamptic sera compared with normal controls and diminished rapidly postpartum to levels equivalent to normal pre- and postpartum serum. These findings are consistent with endothelial cell injury, a process that we believe plays a central role in the pathophysiology of the preeclamptic syndrome. (AM J OssTET GYNECOL 1988;159:1446-51.)
Key words: Preeclampsia, pregnancy, mitogens, endothelial cells, serum growth factors
Preeclampsia occurs in 7% to 10% of pregnancies and is responsible for significant maternal and fetal morbidity. 1 Despite decades of interest and research, the pathogenesis of this disease is still poorly understood. Recent evidence, however, suggests that endothelial cell injury may play an important role in the preeclamptic syndrome. The histopathologic findings of endothelial lesions in renal and umbilical vessels obtained from preeclamptic patients have been recognized.2· 3 Lately, several reports have documented biochemical abnormalities in serum from preeclamptic patients, which support the concept that endothelial cell perturbation and sublethal injury of these cells may contribute to the pathogenesis of preeclampsia. Most of these studies have been indirect and have shown elevated levels of endothelial cell products in serum, for example, fibronectin,4 factor VIII antigen, 5 or platelet and coagulation abnormalities. 1 We have recently From the Departments of Obstetrics, Gynecology, and Reproductive Sciences, Medicine, and Laboratory Medicine, University of California, San Francisco. Supported by National Institutes of Health Grants HD07162, HD22873, and HD24180. Presented in part at the Thirty-fifth Annual Meeting of the Societ_~ For Gynecologic Investigation, Balitmore, Maryland, March 1720, 1988. Reprint requests: Robert N. Taylor, MD, PhD, W. M. Keck Laboratory of Development and Molecular Genetics, Department of Obstetrics, Gynecology. and Reproductive Sciences, M-1489, University of California, San Francisco, San Francisco, CA 9414 30132. 1446
demonstrated directly that serum from preeclamptic women injures endothelial cells in vitro." Advances in the general understanding of endothelial cell function and pathobiology have led to the concept that endothelium has profound effects on the regulation of vascular tone, vascular remodeling, and the activation of platelets and the coagulation cascade after injury. 7 Endothelial cell injury can lead to the secretion of potent mitogens and vasoactive factors through increased production of peptide growth factors by endothelium itself" or by stimulating the release of stored products by activated platelets." This study was undertaken to test the hypotheses that endothelial cell iajury in preeclamptic patients caused increased secretion of mitogenic factors and that growth-promoting activity could be detected in the sera. A bioassay was used to quantify mitogenic activity in these sera by assessing their ability to trigger mitosis in quiescent (G 0 stage) cultured human fibroblasts.
Material and methods Patients. Consecutive patients were recruited from admissions to the Obstetrical Service of the Medical Center at the University of California, San Francisco. All had term, singleton gestations, and the diagnosis of preeclampsia was made on the basis of the following criteria: no prior history of hypertension or renal disease, a rise in blood pressure of at least 30 mm Hg systolic or 15 mm Hg diastolic, or if these were not
Volume 159 Number 6
Serum mitogens in preeclampsia
1447
Table I. Patient data
Gestational age (wk) Mean arterial pressure (mm Hg) Urine protein (mg/di) Uric acid (mg/di) Platelets (Kimi)
Normal
Preeclamptic
(n = 14)
(n = 15)
39.5 ± 1.5
38.5 ± 1.9
84 ± 3.1 0 - trace NDt ND:j:
107 ± 10.7* 30 - 300 5.6 ± 1.8 261 ± 65
Values are expressed as mean ± SD. ND, Not determined.
*p < 0.01. t Normal value at 37 to 40 weeks
= 4.4
± 0.8 mg/di. Kimi.
:j: Normal value (nonpregnant) = 150 to 400
known a blood pressure of at least 140 mm Hg systolic or 90 mm Hg diastolic (manifested on two readings at least 6 hours apart), and proteinuria of ~ 1 + (-30 mg/di) urine protein on a catheterized specimen. Normal control patients had no history of hypertension, were normotensive throughout gestation, and had no proteinuria (Table I). Edema was not used to define preeclampsia for inclusion in this study. Mean arterial pressure was calculated from the average of blood pressure readings taken from admission to delivery (mean arterial pressure = diastolic + [systolic - diastolic]/3). Serum samples. Venous blood samples were collected in early labor and again at 24 to 48 hours post partum in accordance with a protocol approved by the University of California, San Francisco Committee on Human Research. The serum fraction was separated by centrifugation and stored frozen for as long as 4 months at - 70° C before assay. No assays were performed with fresh serum. Cell culture. Human foreskin fibroblasts (passages IO through 20) were obtained from the University of California, San Francisco Cell Culture facility and maintained in culture by means of a growth medium of Dulbecco's modified Eagle's medium with 4.5 gm/L of glucose supplemented with 10% fetal calf serum, penicillin/streptomycin (100 U/ml), and fungizone (500 ng/ml) at 37° C in a humidified atmosphere of 5% C0 2 • Cells were plated at a density of 5 x 10' cells per well in 24-well Falcon dishes, grown to confluence, and then maintained in a quiescent state with serumfree medium for 48 to 72 hours before exposure to patient serum. Serum-free medium was prepared with Dulbecco's modified Eagle's medium but supplemented with 20 mmol/L HEPES, insulin (l µg/ml), transferrin (5 µg/ml), and bovine serum albumin (500 µg/ml, Sigma Chemical Co., St .. Louis, Mo.) in place of fetal calf serum. Thymidine incorporation assay. Mitogenic activity
of serum was determined by measuring the incorporation of tritiated thymidine into nascent deoxyribonucleic acid by fibroblasts as described by DiCorleto and Bowen-Pope.' 0 Briefly, aliquots of patients' sera were added to triplicate wells containing confluent, quiescent, fibroblast cultures to yield a final serum concentration of 2% (final protein concentration ± SD = 1.5 ± 0.4 mg/ml). After 20 hours of incubation, tritiated thymidine (0.25 to 1 µCi/well, 15.7 Ci/mmol, New England Nuclear, Boston, Mass.) was added to each well and labeling was carried out for 4 hours. Thymidine incorporation into 10% trichloroacetic acid-insoluble nucleic acid was determined by scintillation counting. Mitogenic activity was expressed either as total counts per minute oftritiated thymidine incorporated per well of confluent fibroblasts or as percentage of mitogenic stimulation. The latter was defined as (counts per minute incorporated before delivery)/(counts per minute incorporated after delivery) X 100. All values were ex· pressed as the mean ± SE of triplicate assays. Protein determination. Protein concentrations of the serum samples were quantified by the method of Bradford.'' Statistics. Evaluations of statistical significance were performed by means of analysis of variance and the Student t test (two-tailed analysis) where appropriate. Tests with p < 0.01 were considered to reflect significant differences. Results Comparison ofmitogenic activities of pre- and postpartum serum. Initial studies were performed to determine the mitogenic effect of increasing concentrations of serum on human foreskin fibroblasts in culture. There was a dose-dependent increase in tritiated thymidine incorporation from 0.04 to 4 mg/ml of serum protein. No difference in mitogenic potency was seen
1448
Musci et al.
December 1988 Am J Obstet Gynecol
Ci)
~
-...
"C
100000
Cl)
cu
0
...a.
0 0
c
50000
E
a. 0
0
A
.1
10
serum concentration (mg/ml protein)
Ci)
~
-......
"C
100000
Cl)
cu
0
a.
0
0
c E
50000
a.
0
.1
0
B
10
serum concentration (mg/ml protein)
Fig. 1. Mitogenic dose-response of cultured fibroblasts exposed to serum from normal and preeclamptic patients. Quiescent fibroblasts were incubated with paired pre- and postpartum sera from normal and preeclamptic patients as described in the Material and Methods section. Incorporation of tritiated thymidine into fibroblast deoxyribonucleic acid increased in a dose-dependent, logarithmic fashion. Shown are mitogenic activities of prepartum (open box) and postpartum (closed box) sera from normal (A) and preeclamptic (B) patients. There was a significant left shift in the curve for prepartum compared with postpartum preeclamptic serum but not in paired sera from normal pregnancy (p < 0.01).
in paired sera from normal parturients (Fig. 1, A). However, matched pre- and postpartum specimens from preeclamptic patients showed a significant left shift in the dose-response curve in the prepartum spec-
imens compared with postdelivery samples (p < 0.01 for serum concentrations > 1 mg/ ml of protein, Fig. 1, B). These data defined the range of serum protein concentrations that allowed differences to be detected
Serum mitogens in preeclampsia
Volume 159 Number 6
1449
80000
. 60000
a; ~
-...
P<0.01
-
~
II
'tJ Cl)
Q.
40000
-
u .E E Q. u
20000
-
0
...0
~
0
A
normal
preeclamptic
80000 P>0.6 60000
a; ~
-......
-
'tJ
--.---
Cl)
CIS
0
Q.
40000
-
20000
-
--.---
0
u c:
E
Q.
u
B
MgS04
12:1 noMgS04
as
0
normal
preeclamptic
Fig. 2. Mitogenic activity of sera from normal and preeclamptic patients. A, Direct comparison of prepartum sera from preeclamptic and normal patients diluted to a final concentration of 2% demonstrated a 46% increase in mitogenic activity in preeclamptic sera compared with normal sera (p < 0.01). B, Serum collected at 24 to 48 hours' postpartum revealed no difference between the two groups (p > 0.6).
in the subsequent cross-sectional analyses. Experiments comparing a larger number of patients were done at a fixed, final serum concentration of 2% (final serum protein concentration ± SD = l ..?_ ± 0.4 mg/ml). Paired pre- and eostpartum serum samples of 13 preeclamptic and 10 control parturients were compared and normalized with the postpartum value for each patient. Percentage of mitogenic stimulation
Fig. 3. No indirect effect of magnesium sulfate (MgSO,) treatment on the mitogenic activity of sera from preeclamptic patients. Two groups of preeclamptic patients were compared. One group (n = 4) had received intravenous magnesium sulfate before the prepartum blood sample collection. In the other group (n = 4), blood specimens were collected before treatment with magnesium sulfate. The percent mitogenic stimulation (pre-/post- X 100) values were not different (p> 0.6).
(pre-/ post- x 100) was significantly increased in paired preeclamptic sera (mean ± SE = 137% ± 6%) compared with paired sera of normal patients (mean ± SE = 107% ± 6%; p < 0.01). The coefficient of variation of the thymidine incorporation assay was 12%. No changes in serum-induced mitogenic activity were observed even after multiple freeze-thaw cycles. Mitogenic activity in prepartum serum. To determine whether the elevated mitogenic stimulation seen in paired preeclamptic sera was the result of greater prepartum activity, decreased postpartum activity, or both, comparisons of antepartum and postpartum sera were performed. Direct comparison of prepartum sera from 15 preeclamptic and 14 normal patients showed a 46% increase in the mean level of thymidine incorporation of the preeclamptic specimens over the amount in normal specimens (p < 0.01, Fig. 2, A). However, in a separate experiment, direct comparison of postpartum sera collected 24 to 48 hours after delivery revealed no significant difference in mitogenic activity between normal and preeclamptic patients (n = 12. Fig. 2, B). Effect of magnesium sulfate on mitogenic activity. We addressed the possibility that increased mitogenic activity in preeclamptic versus normal sera might be caused by magnesium sulfate therapy administered uniformly to preeclamptic patients at our institution. The direct addition of magnesium sulfate at concentrations of 2 and 6 mg/di (levels comparable with serum levels in treated patients) to control predelivery serum did not affect the incorporation of tritiated thymidine by these samples (p > 0.6, Table II). To determine whether differences in mitogenic activity between patient groups could be attributed to an indirect effect of
1450
Musci et al.
December 1988 Am J Obstet Gynecol
Table II. No direct effect of magnesium sulfate on mitogenic activity of serum* Magnesium sulfate added (mg/di)
Tritiated thymidine incorporated I well (cpm ±SE)
0
2
42,723 ± 3,046
42,540 ± 3,360
6
41,257 ± 2,438
cpm, Counts per minute. *Quiescent human fibroblasts were incubated for 24 hours in the presence of 2% serum from a normal prepartum woman, who was supplemented with 0, 2, or 6 mg/di (final concentration in the serum) of magnesium sulfate). Tritiated thymidine uptake was determined in triplicate as described in the Material and Methods section. Analysis of variance revealed no differences between the three treatments (p > 0.6).
magnesium sulfate treatment, paired pre- and postpartum sera from two groups of preeclamptic patients were compared. One group (l, n = 4) had received intravenous magnesium sulfate before the prepartum blood sample collection. In the other group (2, n = 4), specimens were collected before treatment with magnesium sulfate. The percentages of mitogenic stimulatioh values were not different (group I: mean ± SE= 135% ± 9.5%; group 2; mean± SE= 141% ± 5.9%, p > 0.6; Fig. 3). Comment
In these studies we have shown that sera from preeclamptic women obtained before delivery have increased mitogenic activity compared with normal pregnant women and that this difference is no longer evident within 24 to 48 hours after delivery. It is well known that serum contains a variety of mitogens capable of stimulating quiescent cells in culture to enter the S phase (deoxyribonucleic acid synthesis) of the cell cycle. 12 Most of these mitogens fall into the class of substances known as polypeptide growth factors. Platelets are a rich source of growth factors found in serum, such as platelet-derived growth factor and transforming growth factor-f3. 13 Endothelial cells also synthesize and secrete polypeptides that are mitogenic for mesenchymally derived cells. ' 0 These mitogens and their messenger ribonucleic acids are expressed at high levels by endothelial cells in vitro in response to thrombin and endotoxin and are thought to play a role in fibroblast proliferation during wound healing. 8 · 13 · 14 In vivo endothelial cells initiate platelet aggregation and secretion in response to stimuli such as catheter trauma and hypercholesterolemia" and may also directly release mitogens and vasoactive peptides at these sites. Given these established endothelial cell responses to injury, chronic endothelial cell damage in patients with preeclampsia would be expected to enhance the mitogenic activity in serum from these patients. The rt:sults of this study support such an hypothesis and are consistent with recent findings in our laboratory of a
direct cytotoxic effect of preeclamptic serum on human endothelial cells of the umbilical vein in vitro. Prepartum preeclamptic serum was found to be cytotoxic to endothelial cells in culture, an effect that was markedly diminished postpartum. 6 In the current study, enhanced mitogenic activity was found in prepartum serum from preeclamptic women at a time when its cytotoxic effect was greatest. It remains to be demonstrated whether there is a direct relationship between the cytotoxic effect documented in vitro and the subsequent synthesis and secretion of polypeptide mitogens in vivo. It is intriguing to consider that the widespread vasospasm characteristic of preeclampsia might result from the potent vasoconstrictor effect of platelet-derived growth factor released at sites of endothelial cell injury. 16 At least two interrelated mechanisms could be proposed. The first involves the primary injury of endothelium by a circulating cytotoxic factor(s) produced by placental tissue. 6 This factor(s) may directly stimulate the local production of platelet-derived growth factorlike peptides by sublethally damaging endothelial cells. Alternatively, local activation of platelets, triggered by endothelial cell injury, could lead to release of plateletderived growth factor and other products from platelet o:-granules. Such an hypothesis could explain the clinical findings of vasospasm, coagulopathy, and endothelial lesions in patients with preeclampsia and the growing evidence that platelet activation and turnover are increased in preeclampsia. ' 7 Although the implication of platelet-derived growth factor in this process is attractive because of its combined mitogenic and vasoconstrictive properties, its role remains speculative at this time. Quantitation of the growth-promoting activity of serum in the immediate postpartum period in our study showed a relatively rapid decrease in activity of preeclamptic serum to levels equivalent to normal pre- and postpartum serum. This observation is consistent with the clinical course observed in preeclamptic patients: resolution of the disease begins only after delivery of
Volume 159 Number 6
the products of conception. Removal of the factor(s) responsible for endothelial cell damage at delivery could eliminate the stimulus for secretion of mitogens by endothelium or platelets. For example, plateletderived growth factor has an extremely short half-life in the circulation, which is consistent with the pattern of mitogenic activity observed in our study." Rapid reduction in mitogenic activity postpartum suggests that not only is the source of injury removed at delivery but that the injurious substance itself is rapidly cleared. A model of endothelial cell injury in preeclampsia recently proposed by Friedman 18 implicates prostacyclin in this process. It has been shown that fetoplacental production of prostacyclin is diminished in preeclampsia and that the loss of its cytoprotective effect on endothelial cells may contribute to endothelial injury seen in these patients. In addition, the resulting imbalance in the thromboxane A2 /prostacyclin ratio, which has been documented in preeclampsia, 19 would favor the platelet-aggregating and vasoconstrictor activities of thromboxane A 2 • Because of the short half-lives of these prostaglandins in the circulation, abnormal ratios would correct rapidly once the fetoplacental unit was delivered, and prostacyclin levels (with their cytoprotective effect on the endothelium) would be restored to normal. Thus the rapid disappearance of elevated serum endothelial cytotoxins and mitogens (markers of cell injury) is consistent with this proposed model of endothelial damage, which implicates a placentally derived substance as the injurious factor in preeclampsia. The source, characteristics, and primary or secondary role of this factor(s) are areas of great interest and current investigation. In conclusion, the presence of increased mitogenic activity in prepartum sera from patients with preeclampsia is consistent with increased secretion of mitogenic factors (polypeptide growth factors) by platelets or endothelium and supports our hypothesis that endothelial injury is a major pathogenic factor in preeclampsia. The increased activity seen in preeclamptic sera appears to be neither a direct nor an indirect effect of prophylactic antiseizure therapy with magnesium sulfate. Further work is necessary to characterize and identify the source of this factor(s) and determine whether it is simply a marker of preeclampsia or if it plays a primary role in the pathophysiology of this disorder.
Serum mitogens in preeclampsia
1451
REFERENCES I. Roberts JM. Pregnancy-related hypertension. In: Creasy RK, Resnick R, eds. Maternal-fetal medicine-principles and practice. Philadelphia: WB Saunders, 1984:703-52. 2. Spargo BH, McCartney CP, Winemiller R. Glomerular capillary endotheliosis in toxemia of pregnancy. Arch Pathol 1959;68:593-7. 3. Dadak C, Ulrich W, Sinzinger H. Morphological changes in the umbilical arteries of babies born to preeclamptic mothers: an ultrastructural study. Placenta I 984;5:41926. 4. Stubbs TM, Lazarchick J, Horger EO III. Plasma fibronectin levels in preeclampsia: a possible biochemical marker for vascular endothelial damage. AM J OBSTET GYNECOL 1984;150:885-7. 5. Redman CWG, Beilin LJ, Stirrat GM, et al. Factor VIII consumption in preeclampsia. Lancet 1977; II: 1249-52. 6. Rodgers GM, Taylor RN, Roberts JM. Preeclampsia is associated with a serum factor cytotoxic to human endothelial cells. AM J 0BSTET GY:-.:ECOL 1988; 159:908-14. 7. Jaffe EA. Cell biology of endothelial cells. Hum Pathol 1987;18:234-9. 8. Harlan JM, Thompson PJ, Ross RR, et al. Alpha-thrombin induces release of platelet-derived growth factor-like molecule(s) by cultured human endothelial cells. J Cell Biol 1986;103:1129-33. 9. Reidy MA, Schwartz SM. Arterial endotheliumassessment of in vivo injury. Exp Mol Pa tho I 1984;4 I :41934. 10. DiCorleto PE, Bowen-Pope DF. Cultured endothelial cells produce a platelet-derived growth factor-like protein. Proc Natl Acad Sci USA 1983;80: 1919-23. 11. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 1976; 72:248-54. 12. Watson JD, Hopkins NH, Roberts JW, et al. The Molecular biology of the gene. 4th ed. Menlo Park, Calif.: Benjamin/Cummings, 1987:962-1005. 13. Fox PL, DiCorleto PE. Regulation of production of a platelet-derived growth factor-like protein by cultured bovine aortic endothelial cells. J Cell Physiol 1984; 121: 298-308. 14. Starksen NF, Harsh GR, Gibbs VC, et al. Regulated expression of the platelet-derived growth factor A-chain gene in microvascular endothelial cells. J Biol Chem 1987;262: 14381-4. 15. Ross R, Raines EW, Bowen-Pope DF. The biology of platelet-derived growth factor. Cell 1986;46: 155-69. 16. Berk BC, Alexander RW, Brock TA, et al. Vasoconstriction: a new activity for platelet-derived growth factor. Science I 986;232:87-90. 17. Stubbs TM, Lazarchick J, Van Do rs ten P, et al. Evidence of accelerated platelet production and consumption in nonthrombocytopenic preeclampsia. AM J 0BSTET GYNECOL 1986; 155:263-5. 18. Friedman SA. Preeclampsia: a review of the role of prostaglandins. Obstet Gynecol 1988;71:122-37. 19. Walsh SW. Preeclampsia: an imbalance in placental prostacyclin and thromboxane production. AM J OBSTET GYNECOL 1985; 152:335-40.
Key words: Preeclampsia, pregnancy, mitogens, endothelial cells, serum growth factors
Preeclampsia occurs in 7% to 10% of pregnancies and is responsible for significant maternal and fetal morbidity. 1 Despite decades of interest and research, the pathogenesis of this disease is still poorly understood. Recent evidence, however, suggests that endothelial cell injury may play an important role in the preeclamptic syndrome. The histopathologic findings of endothelial lesions in renal and umbilical vessels obtained from preeclamptic patients have been recognized.2· 3 Lately, several reports have documented biochemical abnormalities in serum from preeclamptic patients, which support the concept that endothelial cell perturbation and sublethal injury of these cells may contribute to the pathogenesis of preeclampsia. Most of these studies have been indirect and have shown elevated levels of endothelial cell products in serum, for example, fibronectin,4 factor VIII antigen, 5 or platelet and coagulation abnormalities. 1 We have recently From the Departments of Obstetrics, Gynecology, and Reproductive Sciences, Medicine, and Laboratory Medicine, University of California, San Francisco. Supported by National Institutes of Health Grants HD07162, HD22873, and HD24180. Presented in part at the Thirty-fifth Annual Meeting of the Societ_~ For Gynecologic Investigation, Balitmore, Maryland, March 1720, 1988. Reprint requests: Robert N. Taylor, MD, PhD, W. M. Keck Laboratory of Development and Molecular Genetics, Department of Obstetrics, Gynecology. and Reproductive Sciences, M-1489, University of California, San Francisco, San Francisco, CA 9414 30132. 1446
demonstrated directly that serum from preeclamptic women injures endothelial cells in vitro." Advances in the general understanding of endothelial cell function and pathobiology have led to the concept that endothelium has profound effects on the regulation of vascular tone, vascular remodeling, and the activation of platelets and the coagulation cascade after injury. 7 Endothelial cell injury can lead to the secretion of potent mitogens and vasoactive factors through increased production of peptide growth factors by endothelium itself" or by stimulating the release of stored products by activated platelets." This study was undertaken to test the hypotheses that endothelial cell iajury in preeclamptic patients caused increased secretion of mitogenic factors and that growth-promoting activity could be detected in the sera. A bioassay was used to quantify mitogenic activity in these sera by assessing their ability to trigger mitosis in quiescent (G 0 stage) cultured human fibroblasts.
Material and methods Patients. Consecutive patients were recruited from admissions to the Obstetrical Service of the Medical Center at the University of California, San Francisco. All had term, singleton gestations, and the diagnosis of preeclampsia was made on the basis of the following criteria: no prior history of hypertension or renal disease, a rise in blood pressure of at least 30 mm Hg systolic or 15 mm Hg diastolic, or if these were not
Volume 159 Number 6
Serum mitogens in preeclampsia
1447
Table I. Patient data
Gestational age (wk) Mean arterial pressure (mm Hg) Urine protein (mg/di) Uric acid (mg/di) Platelets (Kimi)
Normal
Preeclamptic
(n = 14)
(n = 15)
39.5 ± 1.5
38.5 ± 1.9
84 ± 3.1 0 - trace NDt ND:j:
107 ± 10.7* 30 - 300 5.6 ± 1.8 261 ± 65
Values are expressed as mean ± SD. ND, Not determined.
*p < 0.01. t Normal value at 37 to 40 weeks
= 4.4
± 0.8 mg/di. Kimi.
:j: Normal value (nonpregnant) = 150 to 400
known a blood pressure of at least 140 mm Hg systolic or 90 mm Hg diastolic (manifested on two readings at least 6 hours apart), and proteinuria of ~ 1 + (-30 mg/di) urine protein on a catheterized specimen. Normal control patients had no history of hypertension, were normotensive throughout gestation, and had no proteinuria (Table I). Edema was not used to define preeclampsia for inclusion in this study. Mean arterial pressure was calculated from the average of blood pressure readings taken from admission to delivery (mean arterial pressure = diastolic + [systolic - diastolic]/3). Serum samples. Venous blood samples were collected in early labor and again at 24 to 48 hours post partum in accordance with a protocol approved by the University of California, San Francisco Committee on Human Research. The serum fraction was separated by centrifugation and stored frozen for as long as 4 months at - 70° C before assay. No assays were performed with fresh serum. Cell culture. Human foreskin fibroblasts (passages IO through 20) were obtained from the University of California, San Francisco Cell Culture facility and maintained in culture by means of a growth medium of Dulbecco's modified Eagle's medium with 4.5 gm/L of glucose supplemented with 10% fetal calf serum, penicillin/streptomycin (100 U/ml), and fungizone (500 ng/ml) at 37° C in a humidified atmosphere of 5% C0 2 • Cells were plated at a density of 5 x 10' cells per well in 24-well Falcon dishes, grown to confluence, and then maintained in a quiescent state with serumfree medium for 48 to 72 hours before exposure to patient serum. Serum-free medium was prepared with Dulbecco's modified Eagle's medium but supplemented with 20 mmol/L HEPES, insulin (l µg/ml), transferrin (5 µg/ml), and bovine serum albumin (500 µg/ml, Sigma Chemical Co., St .. Louis, Mo.) in place of fetal calf serum. Thymidine incorporation assay. Mitogenic activity
of serum was determined by measuring the incorporation of tritiated thymidine into nascent deoxyribonucleic acid by fibroblasts as described by DiCorleto and Bowen-Pope.' 0 Briefly, aliquots of patients' sera were added to triplicate wells containing confluent, quiescent, fibroblast cultures to yield a final serum concentration of 2% (final protein concentration ± SD = 1.5 ± 0.4 mg/ml). After 20 hours of incubation, tritiated thymidine (0.25 to 1 µCi/well, 15.7 Ci/mmol, New England Nuclear, Boston, Mass.) was added to each well and labeling was carried out for 4 hours. Thymidine incorporation into 10% trichloroacetic acid-insoluble nucleic acid was determined by scintillation counting. Mitogenic activity was expressed either as total counts per minute oftritiated thymidine incorporated per well of confluent fibroblasts or as percentage of mitogenic stimulation. The latter was defined as (counts per minute incorporated before delivery)/(counts per minute incorporated after delivery) X 100. All values were ex· pressed as the mean ± SE of triplicate assays. Protein determination. Protein concentrations of the serum samples were quantified by the method of Bradford.'' Statistics. Evaluations of statistical significance were performed by means of analysis of variance and the Student t test (two-tailed analysis) where appropriate. Tests with p < 0.01 were considered to reflect significant differences. Results Comparison ofmitogenic activities of pre- and postpartum serum. Initial studies were performed to determine the mitogenic effect of increasing concentrations of serum on human foreskin fibroblasts in culture. There was a dose-dependent increase in tritiated thymidine incorporation from 0.04 to 4 mg/ml of serum protein. No difference in mitogenic potency was seen
1448
Musci et al.
December 1988 Am J Obstet Gynecol
Ci)
~
-...
"C
100000
Cl)
cu
0
...a.
0 0
c
50000
E
a. 0
0
A
.1
10
serum concentration (mg/ml protein)
Ci)
~
-......
"C
100000
Cl)
cu
0
a.
0
0
c E
50000
a.
0
.1
0
B
10
serum concentration (mg/ml protein)
Fig. 1. Mitogenic dose-response of cultured fibroblasts exposed to serum from normal and preeclamptic patients. Quiescent fibroblasts were incubated with paired pre- and postpartum sera from normal and preeclamptic patients as described in the Material and Methods section. Incorporation of tritiated thymidine into fibroblast deoxyribonucleic acid increased in a dose-dependent, logarithmic fashion. Shown are mitogenic activities of prepartum (open box) and postpartum (closed box) sera from normal (A) and preeclamptic (B) patients. There was a significant left shift in the curve for prepartum compared with postpartum preeclamptic serum but not in paired sera from normal pregnancy (p < 0.01).
in paired sera from normal parturients (Fig. 1, A). However, matched pre- and postpartum specimens from preeclamptic patients showed a significant left shift in the dose-response curve in the prepartum spec-
imens compared with postdelivery samples (p < 0.01 for serum concentrations > 1 mg/ ml of protein, Fig. 1, B). These data defined the range of serum protein concentrations that allowed differences to be detected
Serum mitogens in preeclampsia
Volume 159 Number 6
1449
80000
. 60000
a; ~
-...
P<0.01
-
~
II
'tJ Cl)
Q.
40000
-
u .E E Q. u
20000
-
0
...0
~
0
A
normal
preeclamptic
80000 P>0.6 60000
a; ~
-......
-
'tJ
--.---
Cl)
CIS
0
Q.
40000
-
20000
-
--.---
0
u c:
E
Q.
u
B
MgS04
12:1 noMgS04
as
0
normal
preeclamptic
Fig. 2. Mitogenic activity of sera from normal and preeclamptic patients. A, Direct comparison of prepartum sera from preeclamptic and normal patients diluted to a final concentration of 2% demonstrated a 46% increase in mitogenic activity in preeclamptic sera compared with normal sera (p < 0.01). B, Serum collected at 24 to 48 hours' postpartum revealed no difference between the two groups (p > 0.6).
in the subsequent cross-sectional analyses. Experiments comparing a larger number of patients were done at a fixed, final serum concentration of 2% (final serum protein concentration ± SD = l ..?_ ± 0.4 mg/ml). Paired pre- and eostpartum serum samples of 13 preeclamptic and 10 control parturients were compared and normalized with the postpartum value for each patient. Percentage of mitogenic stimulation
Fig. 3. No indirect effect of magnesium sulfate (MgSO,) treatment on the mitogenic activity of sera from preeclamptic patients. Two groups of preeclamptic patients were compared. One group (n = 4) had received intravenous magnesium sulfate before the prepartum blood sample collection. In the other group (n = 4), blood specimens were collected before treatment with magnesium sulfate. The percent mitogenic stimulation (pre-/post- X 100) values were not different (p> 0.6).
(pre-/ post- x 100) was significantly increased in paired preeclamptic sera (mean ± SE = 137% ± 6%) compared with paired sera of normal patients (mean ± SE = 107% ± 6%; p < 0.01). The coefficient of variation of the thymidine incorporation assay was 12%. No changes in serum-induced mitogenic activity were observed even after multiple freeze-thaw cycles. Mitogenic activity in prepartum serum. To determine whether the elevated mitogenic stimulation seen in paired preeclamptic sera was the result of greater prepartum activity, decreased postpartum activity, or both, comparisons of antepartum and postpartum sera were performed. Direct comparison of prepartum sera from 15 preeclamptic and 14 normal patients showed a 46% increase in the mean level of thymidine incorporation of the preeclamptic specimens over the amount in normal specimens (p < 0.01, Fig. 2, A). However, in a separate experiment, direct comparison of postpartum sera collected 24 to 48 hours after delivery revealed no significant difference in mitogenic activity between normal and preeclamptic patients (n = 12. Fig. 2, B). Effect of magnesium sulfate on mitogenic activity. We addressed the possibility that increased mitogenic activity in preeclamptic versus normal sera might be caused by magnesium sulfate therapy administered uniformly to preeclamptic patients at our institution. The direct addition of magnesium sulfate at concentrations of 2 and 6 mg/di (levels comparable with serum levels in treated patients) to control predelivery serum did not affect the incorporation of tritiated thymidine by these samples (p > 0.6, Table II). To determine whether differences in mitogenic activity between patient groups could be attributed to an indirect effect of
1450
Musci et al.
December 1988 Am J Obstet Gynecol
Table II. No direct effect of magnesium sulfate on mitogenic activity of serum* Magnesium sulfate added (mg/di)
Tritiated thymidine incorporated I well (cpm ±SE)
0
2
42,723 ± 3,046
42,540 ± 3,360
6
41,257 ± 2,438
cpm, Counts per minute. *Quiescent human fibroblasts were incubated for 24 hours in the presence of 2% serum from a normal prepartum woman, who was supplemented with 0, 2, or 6 mg/di (final concentration in the serum) of magnesium sulfate). Tritiated thymidine uptake was determined in triplicate as described in the Material and Methods section. Analysis of variance revealed no differences between the three treatments (p > 0.6).
magnesium sulfate treatment, paired pre- and postpartum sera from two groups of preeclamptic patients were compared. One group (l, n = 4) had received intravenous magnesium sulfate before the prepartum blood sample collection. In the other group (2, n = 4), specimens were collected before treatment with magnesium sulfate. The percentages of mitogenic stimulatioh values were not different (group I: mean ± SE= 135% ± 9.5%; group 2; mean± SE= 141% ± 5.9%, p > 0.6; Fig. 3). Comment
In these studies we have shown that sera from preeclamptic women obtained before delivery have increased mitogenic activity compared with normal pregnant women and that this difference is no longer evident within 24 to 48 hours after delivery. It is well known that serum contains a variety of mitogens capable of stimulating quiescent cells in culture to enter the S phase (deoxyribonucleic acid synthesis) of the cell cycle. 12 Most of these mitogens fall into the class of substances known as polypeptide growth factors. Platelets are a rich source of growth factors found in serum, such as platelet-derived growth factor and transforming growth factor-f3. 13 Endothelial cells also synthesize and secrete polypeptides that are mitogenic for mesenchymally derived cells. ' 0 These mitogens and their messenger ribonucleic acids are expressed at high levels by endothelial cells in vitro in response to thrombin and endotoxin and are thought to play a role in fibroblast proliferation during wound healing. 8 · 13 · 14 In vivo endothelial cells initiate platelet aggregation and secretion in response to stimuli such as catheter trauma and hypercholesterolemia" and may also directly release mitogens and vasoactive peptides at these sites. Given these established endothelial cell responses to injury, chronic endothelial cell damage in patients with preeclampsia would be expected to enhance the mitogenic activity in serum from these patients. The rt:sults of this study support such an hypothesis and are consistent with recent findings in our laboratory of a
direct cytotoxic effect of preeclamptic serum on human endothelial cells of the umbilical vein in vitro. Prepartum preeclamptic serum was found to be cytotoxic to endothelial cells in culture, an effect that was markedly diminished postpartum. 6 In the current study, enhanced mitogenic activity was found in prepartum serum from preeclamptic women at a time when its cytotoxic effect was greatest. It remains to be demonstrated whether there is a direct relationship between the cytotoxic effect documented in vitro and the subsequent synthesis and secretion of polypeptide mitogens in vivo. It is intriguing to consider that the widespread vasospasm characteristic of preeclampsia might result from the potent vasoconstrictor effect of platelet-derived growth factor released at sites of endothelial cell injury. 16 At least two interrelated mechanisms could be proposed. The first involves the primary injury of endothelium by a circulating cytotoxic factor(s) produced by placental tissue. 6 This factor(s) may directly stimulate the local production of platelet-derived growth factorlike peptides by sublethally damaging endothelial cells. Alternatively, local activation of platelets, triggered by endothelial cell injury, could lead to release of plateletderived growth factor and other products from platelet o:-granules. Such an hypothesis could explain the clinical findings of vasospasm, coagulopathy, and endothelial lesions in patients with preeclampsia and the growing evidence that platelet activation and turnover are increased in preeclampsia. ' 7 Although the implication of platelet-derived growth factor in this process is attractive because of its combined mitogenic and vasoconstrictive properties, its role remains speculative at this time. Quantitation of the growth-promoting activity of serum in the immediate postpartum period in our study showed a relatively rapid decrease in activity of preeclamptic serum to levels equivalent to normal pre- and postpartum serum. This observation is consistent with the clinical course observed in preeclamptic patients: resolution of the disease begins only after delivery of
Volume 159 Number 6
the products of conception. Removal of the factor(s) responsible for endothelial cell damage at delivery could eliminate the stimulus for secretion of mitogens by endothelium or platelets. For example, plateletderived growth factor has an extremely short half-life in the circulation, which is consistent with the pattern of mitogenic activity observed in our study." Rapid reduction in mitogenic activity postpartum suggests that not only is the source of injury removed at delivery but that the injurious substance itself is rapidly cleared. A model of endothelial cell injury in preeclampsia recently proposed by Friedman 18 implicates prostacyclin in this process. It has been shown that fetoplacental production of prostacyclin is diminished in preeclampsia and that the loss of its cytoprotective effect on endothelial cells may contribute to endothelial injury seen in these patients. In addition, the resulting imbalance in the thromboxane A2 /prostacyclin ratio, which has been documented in preeclampsia, 19 would favor the platelet-aggregating and vasoconstrictor activities of thromboxane A 2 • Because of the short half-lives of these prostaglandins in the circulation, abnormal ratios would correct rapidly once the fetoplacental unit was delivered, and prostacyclin levels (with their cytoprotective effect on the endothelium) would be restored to normal. Thus the rapid disappearance of elevated serum endothelial cytotoxins and mitogens (markers of cell injury) is consistent with this proposed model of endothelial damage, which implicates a placentally derived substance as the injurious factor in preeclampsia. The source, characteristics, and primary or secondary role of this factor(s) are areas of great interest and current investigation. In conclusion, the presence of increased mitogenic activity in prepartum sera from patients with preeclampsia is consistent with increased secretion of mitogenic factors (polypeptide growth factors) by platelets or endothelium and supports our hypothesis that endothelial injury is a major pathogenic factor in preeclampsia. The increased activity seen in preeclamptic sera appears to be neither a direct nor an indirect effect of prophylactic antiseizure therapy with magnesium sulfate. Further work is necessary to characterize and identify the source of this factor(s) and determine whether it is simply a marker of preeclampsia or if it plays a primary role in the pathophysiology of this disorder.
Serum mitogens in preeclampsia
1451
REFERENCES I. Roberts JM. Pregnancy-related hypertension. In: Creasy RK, Resnick R, eds. Maternal-fetal medicine-principles and practice. Philadelphia: WB Saunders, 1984:703-52. 2. Spargo BH, McCartney CP, Winemiller R. Glomerular capillary endotheliosis in toxemia of pregnancy. Arch Pathol 1959;68:593-7. 3. Dadak C, Ulrich W, Sinzinger H. Morphological changes in the umbilical arteries of babies born to preeclamptic mothers: an ultrastructural study. Placenta I 984;5:41926. 4. Stubbs TM, Lazarchick J, Horger EO III. Plasma fibronectin levels in preeclampsia: a possible biochemical marker for vascular endothelial damage. AM J OBSTET GYNECOL 1984;150:885-7. 5. Redman CWG, Beilin LJ, Stirrat GM, et al. Factor VIII consumption in preeclampsia. Lancet 1977; II: 1249-52. 6. Rodgers GM, Taylor RN, Roberts JM. Preeclampsia is associated with a serum factor cytotoxic to human endothelial cells. AM J 0BSTET GY:-.:ECOL 1988; 159:908-14. 7. Jaffe EA. Cell biology of endothelial cells. Hum Pathol 1987;18:234-9. 8. Harlan JM, Thompson PJ, Ross RR, et al. Alpha-thrombin induces release of platelet-derived growth factor-like molecule(s) by cultured human endothelial cells. J Cell Biol 1986;103:1129-33. 9. Reidy MA, Schwartz SM. Arterial endotheliumassessment of in vivo injury. Exp Mol Pa tho I 1984;4 I :41934. 10. DiCorleto PE, Bowen-Pope DF. Cultured endothelial cells produce a platelet-derived growth factor-like protein. Proc Natl Acad Sci USA 1983;80: 1919-23. 11. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 1976; 72:248-54. 12. Watson JD, Hopkins NH, Roberts JW, et al. The Molecular biology of the gene. 4th ed. Menlo Park, Calif.: Benjamin/Cummings, 1987:962-1005. 13. Fox PL, DiCorleto PE. Regulation of production of a platelet-derived growth factor-like protein by cultured bovine aortic endothelial cells. J Cell Physiol 1984; 121: 298-308. 14. Starksen NF, Harsh GR, Gibbs VC, et al. Regulated expression of the platelet-derived growth factor A-chain gene in microvascular endothelial cells. J Biol Chem 1987;262: 14381-4. 15. Ross R, Raines EW, Bowen-Pope DF. The biology of platelet-derived growth factor. Cell 1986;46: 155-69. 16. Berk BC, Alexander RW, Brock TA, et al. Vasoconstriction: a new activity for platelet-derived growth factor. Science I 986;232:87-90. 17. Stubbs TM, Lazarchick J, Van Do rs ten P, et al. Evidence of accelerated platelet production and consumption in nonthrombocytopenic preeclampsia. AM J 0BSTET GYNECOL 1986; 155:263-5. 18. Friedman SA. Preeclampsia: a review of the role of prostaglandins. Obstet Gynecol 1988;71:122-37. 19. Walsh SW. Preeclampsia: an imbalance in placental prostacyclin and thromboxane production. AM J OBSTET GYNECOL 1985; 152:335-40.